module 3: overview of fuel and coolant salt chemistry and ...5 module 3 overview of fuel and coolant...

ORNL is managed by UT-Battelle for the US Department of Energy

Module 3: Overview of Fuel and Coolant Salt Chemistry and Thermal HydraulicsPresentation on Molten Salt Reactor Technology by: David Holcomb, Ph.D.Advanced Reactor Systems and Safety GroupReactor and Nuclear Systems Division

Presentation for: US Nuclear Regulatory Commission StaffWashington, DC

Date: November 7–8, 2017

2 Module 3 Overview of Fuel and Coolant Salt Chemistry and Thermal Hydraulics

What are “Molten Salts”?• Salts are ionic compounds formed from a combination

of electronegative and electropositive elements– At elevated temperatures salts liquefy and are termed

“molten salts”

• Halide salts are ionic compounds formed from the combination of a halogen (electronegative) and another electropositive element – commonly, but not exclusively, alkali metals or alkaline earths

– Examples: LiF, BeF2, MgCl2, NaCl (aka table salt), ZrF4, RbF, UF4, UCl3

(left) Solid “Frozen” and (right) Liquid “Molten” 2LiF-BeF2 salt

Halogens

Alkali Metals

Alkaline Earths


Molten Halide Salts Have Attractive Heat Transfer Properties

Coolant(Reactor Concept)

High Working Temperaturea

High Volumetric

Heat Capacityb

Low Primary

Pressurec

Low Reactivity with Air &

Waterd

Coolant&

MaterialsCost

Water (PWR)

Sodium (SFR)

Helium (GCR)

Salt (FHR/MSR) aHigh system working temperature desirable for high efficiency power conversion and process heat applicationsbHigh coolant volumetric heat capacity enables ~constant temperature heat addition / removal (C = 1 – TC/TH ~ Carnot cycles),compact system architectures, and reduces pumping power requirements

cLow primary system pressure reduces cost of primary vessel and piping and reduces energetics of pipe break accidentsdLow reactivity with air and water reduces energetics of pipe break accidents


Molten Salts Are Attractive Coolants for Very High TemperaturesCompared to 20°C water

Fluorides:• ~ 2X density• ~ 1/2X heat capacity• ~ 1–5X viscosity• ~ 2X thermal conductivity• ~ 1X coefficient of expansion as a liquid• Very low vapor pressure

Chlorides:• ~ 1 1/2X density• ~ 1/4X heat capacity• ~ 1 1/2X viscosity• ~ 1X thermal conductivity• ~ 1 1/2X coefficient of expansion as a liquid• Very low vapor pressure


Characteristics of Fuel Salts and Coolant SaltsAre Available from Review Articles• A fuel salt is a molten salt that contains fissile material

– C. F. Baes, Jr., “The Chemistry and Thermodynamics of Molten Salt Reactor Fuels,” Journal of Nuclear Materials, 51 (1974) 149-162

– W. R. Grimes, “Molten Salt Reactor Chemistry,” Nuclear Applications and Technology, 8(2) (1970) 137–155

– B. R. Harder, G. Long, and W. P. Stanaway, “Compatibility and Processing Problems in the Use of Molten Uranium-Alkali Chloride Mixtures as Reactor Fuels,” Nuclear Metallurgy, Metallurgical Society of the American Institute of Mining, Metallurgical and Petroleum Engineers, 15 (1969) 405-32

• Coolant salts are molten salts with advantageous heat transfer properties– D. F. Williams, Assessment of Candidate Molten Salt Coolants for

the NGNP/NHI Heat-Transfer Loop, ORNL/TM-2006/69


Composition of Fuel Salts Are Tailored to Performance Objectives

• Fuel salts consist of a mixture of – Fissile material– Fertile material (if used) – Solvent (diluent)

• Lowers melting point• Decreases power density• Decreases viscosity

– Fissile oxidation prevention material• Preferentially oxidizes to avoid creation of fissile oxide particles due to

contamination

– Fission products (upon use)

ZrF4 UO2 ZrO2+UF4


Fuel Salts Must Integrate Reactor Physics, Heat Transfer, and Material Compatibility

• Reactor physics requirements– Low neutron absorption

• Thermal neutron absorption is of lower importance for fast spectrum reactors

– Radiolytic stability under in-core conditions– Dissolve fissile materials

• Both chloride and fluoride salts are industrially used as heat transfer fluids

– High heat capacity, high boiling point, low thermal conductivity fluids

– Melting point must be below ~525°C– Relatively insensitive to fission products

• Both fluoride and chloride salts, under mildly reducing conditions, are reasonably compatible with high temperature structural alloys and graphite

Source: Grimes, “Molten Salt Chemistry,” Nuclear Applications and Technology 8(2)

(1970) 137-155.


Fuel Salts Have Multiple Subclasses• Thermal spectrum reprocessing optimized fluoride salts

– FLiBe (27LiF-BeF2) solvent provides optimal neutronic performance• Lithium to beryllium ratio selected to minimize melt temperature with acceptable viscosity• High tritium production – need isotopically separated lithium

– NaF-ZrF4 solvent does not require isotopic separation• Much lower tritium production• Higher vapor pressure

– ~1% fissile loading– Fertile loadings vary but are typically much higher (~20%)

• Fast spectrum and thermal spectrum, once-through fuel cycle optimized fluoride salts

– Much higher fissile loading (actinide-rich eutectics)– Adequate fissile material content is a significant design challenge

• Chloride salts– Enables harder neutron spectrum and enhanced breeding– Isotopically separated chlorine preferable - 35Cl from 37Cl

• 35Cl has a moderate capture cross-section (n,) E < 0.1 MeV < E (n,p)

Chlorinenatural isotopic composition

37Cl = 24.23%35Cl = 75.77%

European Fast Spectrum MSR starting fuel

compositionLiF-ThF4-UF4-(TRU)F3 with

77.7-6.7-12.3-3.3 mol%


Fluoride Fuel Salts Have Substantially More Experimental Data Than Chloride Fuel Salts

• Fluoride salts – Two operating molten salt reactors– Multiple in-pile loops– Many capsule tests– Fast-spectrum fluoride salts have much less experience

• Chloride salts – laboratory measurements of physical properties– No in-core testing of fuel salts– Use in pyroprocessing


Thermal Spectrum Fuel Salt Behaves Similarly to Solvent Salt• MSRE nominal fuel mixture was

65 LiF, 29.1 BeF2, 5 ZrF4, 0.9 UF4 (mol %)

• Uranium enriched to 33%• Uranium trifluoride

disproportionates in most molten fluoride solutions

– Large UF4/UF3 ratio prevents disproportionation

• Isotopically pure 7Li - nominally 99.993% at MSRE– Means to limit tritium production due

to large 6Li cross-section

4UF3 3UF4+U0

LiF-BeF2 Phase Diagram

Source: Benes and Konings, “Thermodynamic properties and phase diagrams of fluoride salts

for nuclear applications,” Journal of Fluorine Chemistry, 130, 2009.


Fluoride Fuel Salts Have Limited Solubility forActinide Trifluorides• Fast spectrum systems operate near solubility

limits– Lanthanide trifluorides compete with actinide

trifluorides• CeF3 substantially displaces PuF3

– Log of actinide trifluoride solubility is roughly linear versus inverse temperature

• Monovalent solvent fluorides dissolve much higher levels of actinide trifluorides– Joint solubility of PuF3+UF3 is much less than

individual components up to 600°C– Solubility has strong temperature dependence

• Plate out during transients possible

– Polyvalent fluorides (e.g., ThF4, UF4, or BeF2) substantially reduce solubility Solubility of PuF3 in FLiBe

Source: C. J. Barton, “Solubility of Plutonium Trifluoride in Fused-Alkali Fluoride-Beryllium Fluoride Mixtures” J. Phys. Chem. , Vol. 64, 1960


Fuel Salt Properties Will Be Impacted by Fission Products• Fission products may be gaseous, solid, or dissolved

– Alkaline and alkaline earth fission products (e.g., Cs and Sr) form stable fluorides (or chlorides)

– Semi-noble fission products plate out on metal surfaces• Potential heat load issue following rapid draining

– Noble fission products form suspended clusters that may plate out

• May elect to actively strip gaseous fission products– Lowers the in-core accident source term– Requires cooling fission product traps– Bubble formation and collapse results in reactivity burps

• Fluoride salts have been extensively examined– Reactors, in-pile loops, capsules– Some uncertainty remains - especially about impact of long-term

build up of fission products

• Chloride fuel salts almost entirely untested in core environments– Potential for development of undesirable compounds and phases

“I am pleased, without benefit of rack and thumbscrew, to recant. More realistic calculations based on the single-region ‘reference design’ MSBR heat exchangers indicate that peak afterheat temperatures, while still uncomfortably high, will be much lower than originally anticipated.”

J. R. Tallackson, ORNL-TM-3145


H HeLi Be B C N O F NeNa Mg Al Si P S Cl ArK Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Ga Ge As Se Br KrRb Sr Y Zr Nb Mo Tc Ru Rh Pd Ag Cd In Sn Sb Te I XeCs Ba La‐Lu Hf Ta W Re Os Ir Pt Au Hg Tl Pb Bi Po At RnFr Ra Ac‐Lr

Fission Product Solubility Changes Along Decay Chain

A few elements are very sensitive to redox changes:

Nb behavior changed during MSRE operation after addition of Be°

Transitional (soluble gas soluble) decay example:

137I 137Xe 137Cs

Nb

6 % cumulative 235U fission

yield4-min. half-life

24 sec. half-life

soluble sometimes solubleinsoluble


Cover Gas Handling System Is a Key Element of Any MSR• Distribution of fission products is a central safety issue

– Reduction of fission products in the core limits potential fuel accident source term

– Fission products away from core change decay cooling requirements and radionuclide containment requirements

• Cover gas will inevitably contain some fission products– Aggressive sparging may result in up to 40% of fission products in cover gas (nearly all

of the fission products with gaseous precursors)– Results in substantial heat load in short term fission product trap– Longer term fission product traps contain much lower levels of activity

• Transition from fission product barrier function to waste handling system along carbon beds is conceptually significant

– 85Kr emerging from final stage could be vented

• Some fuel salt fissile components have significant vapor pressures– UCl4 boils at 791°C

• Some solvents vaporize incongruently – ZrF4 sublimes resulting in snow-like deposits in exhaust piping


NRG (Petten) Recently Began Irradiation Tests of Fuel Salt Capsules

• SALIENT program is trilateral collaboration between NRG, JRC, and TUD

• Fluoride salts initially– Chlorides later stage

• Goals– Handling experience– Salt–graphite interaction– Fission product stability / redistribution– Metal particle size distribution

• Longer term– Waste route for spent molten salt fuel– In-pile molten salt loop for the HFR Petten

Cartoon of potential Petten MSR loop

Image provided by NRG; used with permission.


Isotope Separation Is a Significant Issue for Both Fluoride and Chloride MSRs

• Lithium enables optimal reactor physics– Lithium-6 is a large cross-section thermal neutron absorber that

yields tritium– Lithium isotope separation is also necessary for fusion and PWR

chemistry control– Mercury amalgam-based lithium isotope separation was performed

at industrial scale in the 1950s for defense purposes

• Chlorine– Absorption reactions in 35Cl both produces 36Cl (long-lived

radionuclide) and results in a reactivity penalty– Lack of chlorine isotope separation technology was a key element

in US decision in 1956 to pursue thermal breeder MSR


Removing Oxygen Is a Key Technology Requirement for Both Fluorides and Chlorides• Salts containing excess oxygen are much

more corrosive

• Hydrofluorination for fluoride salts– HF is highly corrosive - performed offline– Also removes other electronegative impurities -

sulfur and chlorine– Ammonium hydrofluoride - NH4HF2 alternative

• Carbochlorination for chloride salts –phosgene (COCl2) or carbon tetrachloride used as reactant– MO2 + CCl4 → MCl4 + CO2

• Oxygen can also be removed from some chloride melts by precipitation as aluminum oxide– AlCl3 + UO2 → AlO2 + UCl3

Source: Taube EIR-332; p.156

Group


Tritium is Significant Issue For Lithium-Bearing Salts• Tritium is produced by neutron

reactions with lithium, beryllium, and fluorine as well as being a ternary fission product– Tritium production levels are

similar to HWRs

• Tritium chemical state in salt is determined by redox conditions– TF (oxidizing) or T+ (reducing)

• Above 300°C tritium readily diffuses through structural alloys– Heat exchangers represent

largest surface area for diffusion

• Escape through power cycle is potential route for radionuclide release into environment

0.0001

0.001

0.01

0.1

1

10

100

1000

10000

0.00001 0.0001 0.001 0.01 0.1 1 10 100 1000 10000 100000 1000000 10000000

Cross S

econ

(b)

Energy (eV)

Source: Mays, ORNL/TM-5759


Tritium Mitigation Methods Include Stripping, Blocking, and Trapping

• Largest technical challenge for stripping is the small diffusion of tritium in salt– Necessitates intimate mixing of salt and stripping material

• Gas sparging or spraying in gas space using fine droplets• Turbulent flow (to promote mixing) across large surface area window (e.g.,

double-walled heat exchanger)• Flow through packed bed of absorbers

– Palladium alloys have highest tritium diffusion coefficient• Nickel may be acceptable and is much less expensive

– Carbon traps tritium at operating temperatures - desorbs at high temperatures (peak storage at ~800°C)• Nickel coating carbon improves trapping kinetics• Irradiation damage significantly increases number of traps

– Several lanthanides form stable tritides (e.g., Y or Sm)

• Tritium trapping in coolant salt was demonstrated in NaF-NaBF4(8–92 mol%) at engineering scale for MSBR


Molten Salts Have Attractive Heat Transfer Properties

• Large heat capacity and low viscosity are key properties

Source:Nuclear Engineering Handbook 9-90, D. F. Williams et al., ORNL/TM-2006/12

ρ = densitycp= heat capacityμ = dynamic viscosityβ = volumetric expansion coefficient


Molten Salt Passive Cooling Characteristics are Favorable

• Volumetric expansion with temperature provides buoyancy driving force




Salts Have Sharp Boundary Layer (High Prandtl Number)

• Turbulence is required for effective heat transfer (or tritium stripping)




Salt Viscosity Decreases with Temperature

• Flow increases to hotter regions

• Improves temperature uniformity

Source:D. F. Williams et al., ORNL/TM-2006/12D. F. Williams, ORNL/TM-2006/69


Significant Uncertainty Remains in Fluoride Salt Turbulent Heat Transfer

• Little experimental data with few material combinations and geometries

• Y-axis is a common heat transfer correlation for fully developed turbulent flow in tubes

Source:Yoder, ICAPP 14332, 2014

Nu = Nusselt numberPr = Prandtl numberRe = Reynolds number



Laminar Flow Heat Transfer Also Has Significant Remaining Uncertainty

• Axes selected to enable comparison with prior laminar flow correlations (Seiderand Tate)

• Peclet number is a dimensionless ratio of the thermal energy convected to the fluid to the thermal energy conducted within the fluid


Significant Remaining Uncertainty in Prediction of Conductive / Convective Heat Transfer Ratio

• Plot compares experimental and predicted conductive/convective heat transfer ratios

– Prediction based upon reference Gnielinski correlation - commonly used for heat transfer comparisons



Natural Circulation Heat Transfer Has Significant Remaining Uncertainty

• Product of Grashof and Prandtl number (X-axis) is the Rayleigh number associated with buoyancy-driven flow

– Above critical Rayleigh number heat transfer is primarily convection below primarily conduction

– Y-axis is ratio of convective to conductive heat transfer

Gr = Grashof number

Source:Yoder,ICAPP 14332, 2014


Heat Transfer Uncertainties Affect Operating Margin Calculations

• Material combinations and geometries of interest to MSRs have not been thoroughly characterized in past experiments

• Sources of experimental uncertainty include:– Salt purity and purification during the experiment– Film layers/deposits on heated surfaces– Temperature

• More targeted, controlled experimental data is required to improve the confidence in thermophysical property correlations


Molten Fluorides Are Highly Thermodynamically and Radiolytically Stable

• Salts are combinations of strongly electronegative elements with strongly electropositive metals– Very high bond energies– Negative change in Gibbs free energy (-∆Gf) > 100 kcal/mol-F– Structural metal fluorides have Gibbs free energies at least

20 kcal/mol-F less negative• MSRE graphite and Hastelloy N

exposed to coolant salt was untouched after ~3 years of operation

– Salt radiolysis is overwhelmed by recombination at operating temperatures

Source: ORNL/TM-4174


Thermochemical Stability Drives Both Corrosion and Fissile Solubility

• Increased free chlorine results in larger amounts of dissolved structural alloy chlorides

• Increasing ratios of UCl4/UCl3 restrict acceptable choice of structural alloys

• Use of nickel-based structural alloys restricted to UCl4/UCl3 ratios of roughly 0.003 to 5%

– Smaller amounts of UCl4 results in disproportionation of UCl3

4UCl3 ⇄ UCl4+U0

• Refractory coatings would enable higher UCl4/UCl3ratios Source:

Harder, Long, and Stanaway, Nuclear Metallurgy 15:405-432, 1969.• PuCl3 disproportionation is less favorable than that of UCl3


Phase Diagrams of Chloride Fuel Salts Show Fuel System Design Options

• A ~500°C melt point can be achieved with a range of UCl3 to UCl4 ratios– Systems with higher UCl3

fractions have lower uranium loading

– Systems with higher UCl4fractions are more oxidizing (corrosive)

Source: V. N. Desyatnik et al. At. Energ., 31 [6] 631-633 (1971); Sov. At. Energy (Engl. Transl.), 31 [6] 1423-1424 (1971). Fig. 05953—System NaCl-UCl4-UCl3


Maintaining Mildly Reducing Redox Conditions Key to Enabling Use of Engineering Alloys

Source:Baes, Keiser, ORNL/TM-6002

• Use of a circulating redox buffer provides means to maintain redox condition– Fission changes oxidation state of

salts

• Ratio of U4+/U3+ serves as a measure of the redox potential of the salt– Applicable to both fluoride and

chloride salts– Adding beryllium to FLiBe

• Fluoride salts will likely have an ideal ratio of ~10–100


Fission Process Continuously Alters the Fuel Salt Redox Conditions• When a U or Pu ion fissions, the

available electrons will rearrange on each fission product to satisfy its valence requirements and produce either net oxidizing or reducing conditions in the melt– For 235U (as UF4) four F ions are

released. The fission products require less than four and thus there will be an excess of F ions with net oxidizing conditions

– For 239Pu (as PuF3) three F ions are released. The fission products require more than three and thus there will be a F ion deficit with net reducing conditions

• MSRE periodically added metallic beryllium (strong reducing agent) to maintain UF4/UF3 ratio

Salt Type Fission Product Oxidation

State (Z)Yield (Y) [atoms]

Cl atoms reacted

(Y*Z)

Chl

orid

e Sa

lt (U

Cl 3)

Kr, Xe 0 25 0Rb, Cs 1 19 19Sr, Ba 2 10 20Rare Earths 3 46 138Zr 3 22 66Nb, Mo 0 2 0Te, I 0 6 0Pd, Re, RhAg, Cd 0 61 0

Total Cl atoms reacted out of 300 available 243

Fluo

ride

Salt

(UF 4

)

Br, I -1 1.5 -1.5Kr, Xe 0 60.6 0Rb, Cs 1 0.4 0.4Sr, Ba 2 7.2 14.4Lanthanides, Y 3 53.8 161.4Zr 4 31.8 127.2Nb 0 1.4 0Mo 0 20.1 0Tc 0 5.9 0Ru 0 12.6 0

Total F atoms reacted out of 400 available 301.9

Sources: Baes (fluoride salts), Harder (chloride salts)


Because Chemical Activity in Molten Salts Is Controlled by Melt Composition…• Monovalent salts are “basic” in that they supply fluoride ions

(F-)• Polyvalent salts are “acidic” in that they form complexes

with F−

• Lewis acid/base coordination equilibria are established– ZrF4 + 3F– ↔ ZrF7

3–

– BeF2 + 2F– ↔ BeF42–

• The chemical reactivities of these and other metal ions are higher when they are not sufficiently coordinated with fluoride ions

• In the absence of the extra fluoride ions supplied by LiF component, for example, ZrF4 and BeF2 would be volatile and distill from the system


Fission Products and Contaminants Would Alter Fuel Salt and Cover Gas Properties• Oil leak along MSRE pump shaft resulted in foaming in

pump bowl– Foam overflowed into gaseous waste handling system

• Noble fission products do not dissolve into salt and consequently lack a surface tension inhibition for entering cover gas (i.e., they readily enter the cover gas)

• Contamination particles, solid oxide precipitate, etc., may form a scum layer on the salt surface

Source:Yoder et al., ORNL/TM-2014/499


Fluoride Salts Are Vulnerable to Radiolytic Decomposition at Low Temperatures• Intense radiation creates more free fluorine than is

recombined below ~200°C• Experience with chloride salts is almost nonexistent

– Likely has similar vulnerability as fluoride salts – Pyroprocessing salts and conditions are different from fuel salts

• Free fluorine can react with structural materials resulting in dramatically increased corrosion or converting solid UF4 to gaseous UF6– Origin of the issue with the stored MSRE fuel salt in the 1990s

Source:Haubenreich, ORNL-TM-3144, 1970


Long Term Waste Forms from MSRs Remain Unproven• Primary US work remains “Applied Technology”

• Offgas sorbent could serve as ultimate fission gas disposal medium– Charcoal beds were employed for MSRE

• May be possible to make fluorides more stable by conversion to a fluorophosphate

• Chlorides are currently converted to a salt-cake waste form as part of the ongoing EBR-II processing campaign

• Synthetic rock process developed by ANSTO appears applicable

• Dutch SALIENT project has primary objective to develop final waste form for their test salts


Characteristics of MSRs Derive from the Chemistry and Physics of Halide Salts• Low pressure, high temperature operation• Dissolve useful amounts of fissile material

• Chemically compatible with engineering alloys in mildly reducing environments

• Strong passive safety features– Negative reactivity feedback– Natural circulation-based decay heat removal– Reduced potential for radionuclide release

• Fluoride salts have substantially more experimental data than chloride salts for reactor operations

• Tritium production from lithium-bearing salts can be mitigated by stripping, blocking, and trapping

module 3: overview of fuel and coolant salt chemistry and ...5 module 3 overview of fuel and coolant...

Documents